My made ICC Paper Profiles have less gamut volume than the Paper company's

[cengell]

Hello all, I have been making my own paper profiles with x-rite i1 profiler v1.5 on Win 7 for a few months on a Canon IPF6300 12 color with Canon's ink and have purchased papers from Red River and downloaded there profiles for that same papers for the same Canon printer and on glossy and luster and matte papers, I am about 50,000 to 200,000 smaller gamut volume on many of the Red Rivers papers, I use Chromix Color Think Pro 3.03 on Windows and see that my made profiles are smaller and also noticed most of there's are done on a MAC OS.

I am printing 1728 patches scrambled (I was told you get better profile with the scrambled on) waiting 2-4 days for them to dry fully. Using a X-rite i1 spectro (maybe some of the much more costly Spectro make a better profile or larger gamut?)

I am doing research on the web and I just started reading Andrew Rodney book and Andrew knows allot and hope he will comment in please? I understand at 1728 patches that by printing more patches 2000 to 4096 patches you may not make a larger gamut volume and sometimes less.

I guess I am looking for some ideas with i1 Profiler or some other software for Windows. I understand that it's not always about the numbers like gamut volume, what counts is what the final image on paper looks like but I want to learn what I am doing wrong or right or what I can do to get better gamut volume on my papers.

Thank you for any help.

Christopher

[Scott Martin]

Are you using the same media setting as Red River's are made with? What Perceptual rending options are you using in i1P? [Tip: crank up the Saturation to 45]

How do your prints look? If you print an evaluation image with color ramps etc using the Perceptual and RelCol intents with both profiles how do they compare? Are the results visually consistent with the gamut rendering results you're seeing in ColorThink? [feel free to download my eval image from http://www.on-sight.com/downloads/ ]

If you're using the same media settings, I'll bet your profiles visually look better than theirs counter to what the gamut rendering suggests. That would be common - profiles made with different software often render differently. The bottom line is that you can't trust the gamut renderings when comparing profiles from different manufacturers and should instead trust what you see visually. More elaborate, empirical testing would involve measuring CMYKRGB color patches printed with each profile.

Comparing gamut rendering with profiles that were made differently is a common trap that many fall into. It can lead to false conclusions quickly! Real world print testing and visual analysis is essential...

[cengell]

Hello Scott, thank you for your help I will try that. Also I am doing RGB profiles and using the same settings as Red River says too, of course I really don't know what they did 100%. I really don't like the RGB profiles because you really can't set a ink limit (and not sure if that would affect gamut volume ?)

I would think I should at least get very close to what they are getting or what Canson or Innova and anyone else, as they are doing a RGB profile but with a more costly specrto and software? I think you were also on the i1 beta team with v1.0 as I recall seeing your name on the the webnars during the beta testing.

I will try to increase sat to 45 I see 50 is max, also I am making a V2 not V4 ICC and did not really know what the slider for smoothness had a effect in doing to the ICC any ideas?

I also played around with the optimization of the profiles and really got no gamut improvement or gamut volume, so I stopped trying that option.

Thank you Scott
Christopher   

[Scott Martin]

I really don't like the RGB profiles because you really can't set a ink limit (and not sure if that would affect gamut volume ?)

The ink limiting is handled my the media setting and the separation parameters are handled on-printer. I think you should like RGB profiles - they are so simple and fantastic!

I would think I should at least get very close to what they are getting or what Canson or Innova and anyone else, as they are doing a RGB profile but with a more costly specrto and software? I think you were also on the i1 beta team with v1.0 as I recall seeing your name on the the webnars during the beta testing.

Thanks for noticing. I was the first and only beta tester starting in 2005 at the beginning of the Prism engine development long before the Munki and i1P products came to market. Has been a long road!

I will try to increase sat to 45 I see 50 is max, also I am making a V2 not V4 ICC and did not really know what the slider for smoothness had a effect in doing to the ICC any ideas? I also played around with the optimization of the profiles and really got no gamut improvement or gamut volume, so I stopped trying that option.  

Long story short - set smoothness to 70, the spec won't make a visual difference and optimization won't help if you're starting with as many patches as you are.

[digitaldog]

Depending on what product built the two sets of profiles, I wouldn't expect them to be the same so don't worry about it. There is very little you can do wrong when creating RGB profiles with your package.

[cengell]

Thank you Scott! I really do like RGB profiles because they are much easier to make than CMYK not sure if they are a good or have as much or more gamut volume? But your right that choosing the media profile and maybe at random, but I see you have a test image for "Onsight Media Selection Image".

But on the Canon IPF6300 when I print a i1 target I set the Canon driver to no color management correct? So this method with that image won't work correctly? But would work if I use Red River's profile as they tell you which paper to set the Canon paper to.

Right now I got some Epson Cold Press Natural & Bright paper and really don't know where to start since Epson don't have a profile for Canon LOL. So Then I could use your test image to see which setting say of 4 choices is the best then use no color management correct or do I have that wrong?

From now on I will set Sat to 45 and smoothness to 70 as you suggest and see if I get a larger gamut volume and or a better image printed. I won't bother to do any more optimization as I did not see any improvement in Color Think.

Thank you very much Scott and any other thoughts we can email privately if you like.

Christopher

[cengell]

Thank you Andrew, I understand its like apples and grapes are both fruit but you can't compare ?

I was hopping to making the ICC profiles I would get a larger gamut volume or the same or close, but like I said before they may not be using a i1 spectro and using a better unit. 

I just want to learn how to produce the best ICC profile and hope you might have so tips and tricks, I read you got a different software package that you are liking over i1? Do you get a larger gamut or more colors you need?

Thank everyone for your advise I really appreciate you and Scott's input!

Christopher

[Scott Martin]

But on the Canon IPF6300 when I print a i1 target I set the Canon driver to no color management correct?

Yes, or even better, use the Photoshop iPF printing plug-in and set the profile to "none". The plug-in has a more reliable 'no color management' path that never fails.

So this method with that image won't work correctly? But would work if I use Red River's profile as they tell you which paper to set the Canon paper to. 

No, the idea is to analyze to native output *without* a profile.

Right now I got some Epson Cold Press Natural & Bright paper and really don't know where to start since Epson don't have a profile for Canon LOL. So Then I could use your test image to see which setting say of 4 choices is the best then use no color management correct or do I have that wrong?

Right, make a bunch of prints without profiles, pick the one that appears to be the best (takes an eye for this and some experience) then proceed by printing your profiling target and making the profile. That's the idea at least - there are more advanced procedures that we won't talk about here but you get the idea.

For Epson Cold Press on the 6300 I'd recommend the "Premium Matte Paper" media setting. I think you'll be happy with that and it has a Monochrome Photos mode and tiny margins.

From now on I will set Sat to 45 and smoothness to 70 as you suggest and see if I get a larger gamut volume and or a better image printed. I won't bother to do any more optimization as I did not see any improvement in Color Think.

Right and don't even bother comparing the color gamut of your profile to theirs in ColorThink - make prints instead and compare those! Of course, you can compare different profiles that you've made yourself using the same device and software - this is a fair comparison that you can judge in ColorThink.
[/quote]

[cengell]

Thank you Scott! I will do that, I was not 100% sure that in Photoshop CS6 the Canon plugin worked right, now I do! Do you know if that Canon plugin is working in Photoshop CC? I really hate Adobe's ACPU, just horrible as they could have done so much better in my opinion.

I now have a direction to start working on, again thank you and I will still look other profiles but just not be disappointed when I see mine 

I may email you directly if that's alright with you as I also am working with QTR and a 6 grey inkset + 2 LC & LM to change the tone from warm to cool.

Christopher

[digitaldog]

I was hopping to making the ICC profiles I would get a larger gamut volume or the same or close, but like I said before they may not be using a i1 spectro and using a better unit. 

I'm not sure that a different Spectrophotometer would make much difference. To be honest, I've never done a side by side test, using the same target on say an i1 and my iSis. It would still not be totally apples to apples due to the number of samples taken per patch and the patch size does play a role here (with both units).

I just want to learn how to produce the best ICC profile and hope you might have so tips and tricks, I read you got a different software package that you are liking over i1? Do you get a larger gamut or more colors you need?

I'm now using ColorLogic Copra for all CMYK profiles but haven't really done a lot of work with it on RGB. The ColorLogic color engine is producing better output than i1P with the same data, a lot of averages of ECI's for output to press. That is apples to apples testing in this case. We had to hammer them a bit on the issue of blues being mapped from sRGB to our CMYK output going magenta in a couple areas and once we showed them our issues, they fixed the color engine in a few days. Impressive! We're not concerned with gamut volume and to be honest, I haven't even looked at the differences in ColorThink, might do so. It is what happens when ink or toner hits the paper we are more concerned with and Copra is doing a better job than i1P as far as I'm concerned. X-rite needs competition big time and they could really learn to speed up their software fixes like the boys at ColorLogic have done.

[digitaldog]

Just looked at two profiles, built with identical data but one in i1P, one in Copra. Again the Copra profile is producing visually superior output with a large group of real (tough) and synthetic images designed test profiles and their color engines.

The i1P profile has a gamut volume of 440,331. The better profile from Copra has a gamut volume of 440,231. Not much of a difference in the numbers, but Copra is lower but produces better output. So, I'm not very interested in gamut volume numbers, at least in this apple to apple test. The output tells a far more compelling story!

[Scott Martin]

The i1P profile has a gamut volume of 440,331. The better profile from Copra has a gamut volume of 440,231. Not much of a difference in the numbers, but Copra is lower but produces better output. So, I'm not very interested in gamut volume numbers, at least in this apple to apple test. The output tells a far more compelling story!

Yep, I've seen this over and over again where the gamut volume numbers suggest the opposite conclusion from what you would conclude from visual analysis. Gotta make prints!

[digitaldog]

Yep, I've seen this over and over again where the gamut volume numbers suggest the opposite conclusion from what you would conclude from visual analysis. Gotta make prints!

While we have some controls to up (boost) saturation in a perceptual table, the effects upon some images can be more detrimental that useful. Even the fine Roman 16 images can illustrate this. For example, the Blue image has a nice gradient in the bkgnd that really shows off issues in blues shifting magenta, blues inverting (moving along going light to dark, but the going dark then back the other way). Also, depending on the workflow, converting from sRGB versus Adobe RGB or even the original color space of the Roman's can show issues in a profile. If you can always control the source RGB working space for a conversion, that's easier. But sometimes I run into workflows where one is going to CMYK from sRGB while in other cases not.

In the end, making good old RGB profiles and controlling the source color space is so much easier. CMYK really is the devils color space <g>.

[cengell]

Hello Andrew, thanks for that info. Do you know of a free "synthetic images" designed test profiles? I know you have a test printer image as I am a see proof on how well my made ICC profiles are doing? I now understand that the spectro in not the most important part the software is and would I gain or get a better ICC profile by doing CMYK instead of RGB I understand there's more to do but will the final print 1 with a RGB and 1 with same image look different?

I think I recall you at PMA years ago and you had some prints that showed the color spaces sRGB & Adobe 1998 RGB?

I am trying basICColor Print CrossXcolor 1.9.55 engine and I read that it's very close to Copra?

Do you know if you choose a RGB profile Epson or Canon and you use that for the ink limit how much that will affect the ICC profile you make in i1?

Again thank you Andrew

Christopher

[cengell]

Great Scott, I get it. I will first use the finial print and then look at my made ICC profiles. Have you made a RGB and a CMYK profile and printed and compared the 2 prints and did you see any improvement with the RGB or CMYK?

Do you think RGB or CMYK is better getting the most you can for ink & paper?

I want to push my profiles to the maximum possible without spending $$$ like Copra cost unless it's a night and day improvement?

So should I set i1 from the 45 to say 30 to somewhat increase the gamut but not go over and cause issues in prints in the shadows like both of you have seen?

Oh by the way for years I have been to your site and it's great and is Andrew's site.

Thank you Scott

Christopher

[digitaldog]

Hello Andrew, thanks for that info. Do you know of a free "synthetic images" designed test profiles?

One of the best free one's was from Bill Atkinsion (we lovely call it 'Bill's Balls') but the actual name is Twenty-Eight Balls. I'm not sure where you'll find it as Bill seemed have pulled the plug on his site but I recall someone around these parts finding it and uploading it somewhere.

Update: see this, http://www.luminous-landscape.com/forum/index.php?topic=79799.0

[digitaldog]

Do you think RGB or CMYK is better getting the most you can for ink & paper?

This isn't a choice. IF the driver accepts RGB data, that's what I send. When I'm making press profiles, we have to feed that device CMYK and hence that's how we profile it.

[cengell]

Hello Andrew, I see but I guess my question is I have a Canon IPF6300 12 color printer and has a very good gamut and I want to push it to the maximum limits and I know that inkjet paper has allot to do about gamut volume, but taking a high Dmax and very good gamut paper, and can use a CMYK profile in a RIP I just wanted to know if I would get a larger gamut (as you called highest resolution) I can get with that paper and inkset, so on those factors would a RGB or CMYK produce and bigger gamut?


Apples to Apples?

Also is the Gamut volume are numbers of how many colors are capable of seeing in a ICC with the right colorspace, paper and inkset so why can't you compare same paper, printer and inkset by the gamut volume with other made ICC profiles? I am trying to wrap my head around other than some + & - values I would think it's a accurate way to compare?

Thank you very much, I am watching your videos on your site right now, and you suggest I watch to better understand what I am asking?

Christopher

[digitaldog]

Send the Canon RGB!

[Scott Martin]

Hello Andrew, I see but I guess my question is I have a Canon IPF6300 12 color printer and has a very good gamut and I want to push it to the maximum limits and I know that inkjet paper has allot to do about gamut volume, but taking a high Dmax and very good gamut paper, and can use a CMYK profile in a RIP I just wanted to know if I would get a larger gamut (as you called highest resolution) I can get with that paper and inkset, so on those factors would a RGB or CMYK produce and bigger gamut?

Done it and you can't beat the driver's gamut. Don't make it harder than it needs to be - just make RGB profiles as discussed using the driver and I can't tell you - you won't beat the quality. Things were differernt 10 years ago but these days the answer is the driver!